Exp Brain Res (2009) 199:117–126

DOI 10.1007/s00221-009-1979-y

RESEARCH ARTICLE

Music performance anxiety in skilled pianists: eVects of social-evaluative performance situation on subjective, autonomic, and electromyographic reactions

Michiko Yoshie · Kazutoshi Kudo · Takayuki Murakoshi · Tatsuyuki Ohtsuki

Received: 5 June 2009 / Accepted: 5 August 2009 / Published online: 22 August 2009© Springer-Verlag 2009

Abstract Music performance anxiety (MPA), or stagefright in music performance, is a serious problem for manymusicians, because performance impairment accompaniedby MPA can threaten their career. The present study soughtto clarify on how a social-evaluative performance situationaVects subjective, autonomic, and motor stress responses inpianists. Measurements of subjective state anxiety, heartrate (HR), sweat rate (SR), and electromyographic (EMG)activity of upper extremity muscles were obtained while 18skilled pianists performed a solo piano piece(s) of theirchoice under stressful (competition) and non-stressful(rehearsal) conditions. Participants reported greater anxietyin the competition condition, which conWrmed the eVective-ness of stress manipulation. The HR and SR considerablyincreased from the rehearsal to competition conditionreXecting the activation of sympathetic division of the auto-nomic nervous system. Furthermore, participants showedhigher levels of the EMG magnitude of proximal muscles(biceps brachii and upper trapezius) and the co-contractionof antagonistic muscles in the forearm (extensor digitorumcommunis and Xexor digitorum superWcialis) in the compe-tition condition. Although these responses can be inter-preted as integral components of an adaptive biologicalsystem that creates a state of motor readiness in an unstableor unpredictable environment, they can adversely inXuence

pianists by disrupting their Wne motor control on stage andby increasing the risk of playing-related musculoskeletaldisorders.

Keywords Psychological stress · Motor control · Music performance anxiety · Pianist · Electromyography · Autonomic nervous system

Introduction

Music performance anxiety (MPA), more commonlyreferred to as stage fright, has always been aZicting musi-cians all around the world. In a large survey involving2,212 musicians, MPA was found to be the most frequentnon-musculoskeletal medical problem among the respon-dents, with more than 20% using beta-blockers before sig-niWcant public performances (Fishbein et al. 1988). On topof its prevalence, MPA can even threaten a musician’scareer, because an excessive level of MPA sometimes leadsto actual impairment of performance skills (Yoshie et al.2008b, 2009).

So far, research has examined the contribution of situa-tional factors to the arousal of MPA (Papageorgi et al.2007; Wilson 1997). The Wndings suggested that perfor-mance circumstances leading to a strong sense of exposure,including solo against group performance (Cox andKenardy 1993), public performance against practice(Fredrikson and Gunnarsson 1992; LeBlanc et al. 1997),and evaluative against non-evaluative performance (Abeland Larkin 1990; Brotons 1994; Craske and Craig 1984;Hamann and Sobaje 1983; Yoshie et al. 2008a) are likely toincrease MPA and associated physiological arousal. Thus,it seems reasonable to suppose that the combination ofthese situational factors will induce the greatest levels of

M. Yoshie (&) · K. Kudo · T. Murakoshi · T. OhtsukiDepartment of Life Sciences, Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba,Meguro-ku, Tokyo 153-8902, Japane-mail: cc087717@mail.ecc.u-tokyo.ac.jp

M. YoshieJapan Society for the Promotion of Science, 8 Ichibancho, Chiyoda-ku, Tokyo 102-8472, Japan

123

118 Exp Brain Res (2009) 199:117–126

MPA. Furthermore, these studies normally used an experi-mentally devised performing situation where participants’performance outcome would never aVect their future careerprospects. In a real musical world, however, every singleperformance in public can either boost or wreck musicians’career, which causes them higher levels of psychologicalstress. In the present study, therefore, we employed anactual piano competition to observe the stress responses ofmusicians performing in a realistic environment.

Some of the previous MPA studies (Abel and Larkin1990; Brotons 1994; Craske and Craig 1984; Fredriksonand Gunnarsson 1992; LeBlanc et al. 1997) have eVectivelydemonstrated the changes in physiological arousal in stress-ful performances mostly through the measurements of heartrate (HR). However, the relationship between physiologicalarousal and performance has been seen as equivocal(Yoshie et al. 2009), with a variety of predictions rangingfrom the inverted-U hypothesis (Yerkes and Dodson 1908),zones of optimal functioning (ZOF) model (Hanin 1978;Yoshie et al. 2008b) to catastrophe theory (Hardy and Par-Wtt 1991). To better understand the eVects of MPA on per-formance quality, it would be highly beneWcial to examinethe motor process mediating the arousal–performance rela-tionship.

Recent studies have accumulated evidence indicatingthat the induction of psychological stress or negative emo-tions leads to activation of the motor system. Using a func-tional magnetic resonance imaging (fMRI), Butler et al.(2007) showed elevated activation of dorsal basal gangliain response to an experimentally induced state of consciousfear. Transcranial magnetic stimulation studies (Baumgart-ner et al. 2007; Hajcak et al. 2007; Schutter et al. 2008)have also demonstrated enhanced motor-evoked potentialsduring the presentation of fearful or unpleasant emotionalstimuli indicating increased corticospinal motor tract excit-ability associated with negative emotions. These phenom-ena may be understood as representing a state of motorreadiness to facilitate defensive behaviors under a potentialthreat to survival, which has traditionally been known asthe Wght-or-Xight response (Cannon 1915).

These Wndings have also been conWrmed in experimentscombining electromyography (EMG) with simple motortasks. Such studies consistently reported increased EMGactivity in the upper extremity muscles associated with cog-nitive or emotional stress using a wide range of tasks suchas Wnger tapping (Bloemsaat et al. 2005), Wnger elevation(Finsen et al. 2001), computer work (van Galen et al. 2002;Visser et al. 2004; Wahlström et al. 2002, 2003), hand andshoulder exertions (Au and Keir 2007), and ball stroking(Matsumoto et al. 2001). Predictably, elevated muscleactivity under stress has often been accompanied by greaterforce production, as reXected in the increases of isometrickey-pressing force (Bloemsaat et al. 2005), grip- and click-

forces applied on a computer mouse (Visser et al. 2004;Wahlström et al. 2002), maximal isometric force producedby the wrist and Wnger extensor muscles (Coombes et al.2006), pinch grip force (Coombes et al. 2008), and speed ofa stroked ball (Matsumoto et al. 2001). Interestingly, ourrecent experiment (Yoshie et al. 2008a) showed that theseWndings hold for a piano-playing task as well. In the exper-iment, we attempted to manipulate participants’ stress lev-els by simply scoring their arpeggio performance in alaboratory setting, and the resulting increase in state anxietycoincided with elevated EMG activity in the upper extrem-ity muscles and stronger keystroke force. Because strikingthe keys with appropriate force levels is a fundamental skillfor pianists, increased muscle activity stemming from over-all activation of the motor system might account for the lossof Wne motor control in stressful performance situations. Inthe present study, we examined whether such EMG activa-tion is observed also during performances of classical pianoworks in a real competition. We hypothesized that partici-pants would respond to the social-evaluative stressor withhigher muscle activity and less muscle relaxation.

Psychological stress can aVect not only levels of EMGactivity, but also its coordination patterns. Research hasindicated that stress or anxiety increases the co-contractionlevels of antagonistic muscles in the upper extremity(Meulenbroek et al. 2005; van Galen et al. 2002; Weinbergand Hunt 1976). Normally, the levels of co-contraction andjoint stiVness become lower as motor learning proceeds andthe contribution of feedforward component increases (Fujiiet al. 2009a, b; Furuya and Kinoshita 2008; Osu et al.2002). When put under psychological stress, however, theCNS may be required to heighten joint stiVness through co-contraction to maintain movement accuracy in the face ofdeteriorated signal-to-noise ratio in the motor system(Gribble et al. 2003; van Galen et al. 2002; van Gemmert andvan Galen 1997). In this regard, increased co-contractionlevels under stress may be interpreted as a strategic meansto adapt muscle activity to central information processingdemands at the expense of physiological eYciency. Basedon these Wndings, we hypothesized that the co-contractionlevel of antagonistic muscles in the forearm would also behigher in pianists executing over-learned movements underintense social-evaluative stress.

Methods

Participants

We recruited pianists for participation in the competitionexperiment by distributing application forms and informa-tional brochures describing qualiWcation requirements topotential applicants at music schools and piano clubs.

123

Exp Brain Res (2009) 199:117–126 119

Eventually, 18 highly trained pianists (7 men and 11women, mean age § SD = 26.7 § 6.3 years) with a meanof 20.4 (SD = 6.2) years of playing experience participatedin the experiment. Written informed consent was obtainedfrom all the participants, and the study was approved by theEthics Committee of the Graduate School of Arts andSciences, the University of Tokyo.

Experimental task and conditions

We asked participants to perform a solo piano piece(s) oftheir choice requiring considerable playing skills, such aspiano sonatas of Beethoven, ProkoWev, and Rachmaninov.Participants played the whole piece(s) or the Wrst part of thepiece (mean duration § SD = 6.5 § 1.4 min) from memoryon an acoustic grand piano.

Participants attended two sessions on two separate days:the rehearsal and competition. In both conditions, weinstructed participants not to drink caVeinated or alcoholicbeverages within 24 h before their performance. In therehearsal condition, each participant individually visited asmall music practice room, and played the piece(s) on aYamaha C3 grand piano (Yamaha, Hamamatsu, Japan)with no experimenter present in the room. In the competi-tion condition, we held an actual piano competition in aconcert hall, and participants performed the piece(s) on aSteinway D-274 full concert grand piano (Steinway &Sons, NY, USA) in front of a large audience(N = 45.6 § 3.3) and Wve professional judges (four profes-sional concert pianists and a musicologist). After each per-formance, the performer received warm applause from theaudience. In the awarding ceremony following the competi-tion, a judge gave the certiWcates of commendation andcash rewards of 20,000, 10,000, and 5,000 JPY (»200, 100,and 50 USD) to the recipients of the Wrst, second, and thirdprize, respectively. In both the practice room and concerthall, temperature and humidity were carefully controlled.

Data acquisition and analysis

Performance measures

In both conditions, participants’ performances were digi-tally recorded at 44.1 kHz with two microphones, whichwere placed above the piano and connected to a handyaudio recorder H4 (Zoom, Tokyo, Japan). The total dura-tion of each performance was calculated based on the audiorecordings. With respect to the rehearsal condition, the Wvejudges evaluated the recorded performances. In the compe-tition condition, participants were rendered invisible to thejudges so that visual information would not aVect theirevaluation, and the jury directly evaluated participants’ liveperformances. The jury scored performances on ten items,

with four items concerning technical aspects (i.e., accuracy,technical dexterity, tempo and rhythm, and memory) andsix items concerning artistic aspects (i.e., artistry, interpre-tation, expressiveness, structural strength, melodic and har-monic balance, and tone quality). The scale ranged from 1(poor) to 10 (excellent), leading to the total score rangingfrom 10 to 100. To examine the inter- and intra-rater reli-ability of performance evaluation scores, we computedintraclass correlation coeYcients (ICC) for the rankings oftotal scores. The performance scores demonstrated substan-tial levels of inter-rater reliability, with ICC (2, 5) of 0.80and 0.85 (Ps < 0.001) for the rehearsal and competitionconditions, respectively. To conWrm that the jury was capa-ble of consistently evaluating recorded performances andlive performances, we asked three members of the jury torate recorded performances of the competition conditiononce again well after the experiment. Then, we computedICC (1, 1) for each judge using his/her score rankingsdetermined for recorded and live performances of the com-petition condition. The ICC (1, 1) for the three judges were0.77, 0.76, and 0.70 (Ps < 0.001) showing suYcient levelsof intra-rater reliability. Thus, we used the total scores thatwere averaged across the judges to compare performancequality in the competition condition with that in therehearsal condition. We also calculated the mean scores ofthe technical and artistic items separately (technical andartistic scores).

Self-reported measure

The level of state anxiety was assessed just prior to eachparticipant’s performance with the visual analog moodscale (VAS), which is capable of measuring emotionalstates in a quick, reliable, and relatively sensitive manner(Cella and Perry 1986). The VAS is a 100-mm continuousscale ranging from 0 (not anxious at all, the left end) to 100(extremely anxious, the right end). We asked participants toplace a vertical line bisecting the 100-mm line to indicatethe perceived level of anxiety at the moment. The VAS datawere quantiWed by measuring the distance between the leftend and the vertical line (mm).

Autonomic measures

Beat-to-beat HR was monitored continuously throughoutexperimental sessions using a wireless signal transmissiondevice with electrocardiogram precision (Polar S810i byPolar Electro Oy, Kempele, Finland) at a sampling rate of1,000 Hz. The receiver unit of the HR monitor, whichresembled a large wrist watch, was hung from an adjustablebelt worn round the waist in order not to disturb the per-forming pianists. The consecutive inter-beat (RR) intervalswere automatically stored on the receiver. The RR data

123

120 Exp Brain Res (2009) 199:117–126

were carefully edited by visual inspection and eliminationof measurement artifacts. We computed the mean HR(bpm) during performance (i.e., from the onset of the Wrstnote to the end of last note) based on the corrected data.

During each piano performance, we also gauged theamount of sweat evaporated from the sole of the left footwith a sweat rate (SR) meter (TS100 by Techno Science,Tokyo, Japan) utilizing the ventilated capsule technique.The SR data were sampled at 1,000 Hz with an MP150 dataacquisition system (Biopac Systems, Inc., CA, USA) inter-faced with a personal computer. We calculated the meanSR amplitude (mg/min/cm2) for each condition. Figure 1a,b shows the typical examples of HR and SR time series,respectively.

EMG measures

During each piano performance, surface EMG activity wasrecorded from the left extensor digitorum communis (ED),Xexor digitorum superWcialis (FD), biceps brachii (BB),and upper trapezius (TR) muscles. Bipolar Ag/AgCl dis-posable electrodes (10 mm diameter) were placed at theestimated motor point of each target muscle with a 20-mmcenter-to-center distance. At each electrode position, theskin was cleaned using alcohol to reduce skin resistance.During EMG recording, special care was taken to preventthe connecting cables from disturbing the performing pia-nist. The EMG signals were ampliWed 500 times and sam-pled at 1,000 Hz together with SR using the MP150 system.EMG signals were Wrst notch Wltered to remove the 50 Hz(local power line frequency) interference. Then, the signals

were high-pass Wltered at 20 Hz, full-wave rectiWed, andlow-pass Wltered at 50 Hz. Subsequently, baseline noise(i.e., mean resting EMG amplitude of pre- and post-perfor-mance phases) was removed from the signals according toKudo and Ohtsuki (1998).

To normalize these EMG data for each muscle for eachparticipant, maximal voluntary contraction (MVC) datawere obtained by asking participants to perform maximalisometric force production (5 s) after their piano perfor-mance in each condition. Participants were verbally encour-aged to achieve maximal force at designated joint angles.During MVC tasks for the ED and FD muscles, the wristjoint was kept at 180°, and for the BB muscle, the elbowjoint was kept at 90°. In these trials, an assisting personapplied the highest possible load to help participants pro-duce maximal Xexion or extension force. As for the TRmuscle, we adopted the sustained shoulder shrug as theMVC task. An MVC value was determined as the highestmean EMG amplitude observed during the MVC task,which was obtained with a 300 (for the TR muscle) or1,000 ms (for the other three muscles) window moving insteps of 1 ms. EMG signals recorded during performancewere normalized relative to these MVC values (Fig. 2a).

Prior to EMG analyses, we selected appropriate EMGdata by applying systematic criteria. We Wrst performed theSmirnov–Grubbs tests (P < 0.05, with Bonferroni correc-tion) on the percent changes in the mean resting EMGamplitude from the pre- to post-performance phase toexclude EMG data with substantial changes in baselinenoise level (probably due to the detachment of electrodes).In the same way, we excluded EMG data with substantialchanges in the MVC values from the rehearsal to competi-tion condition. The selection process required us to excludethe EMG data of three participants for the TR muscle andthe data of two participants for the other three muscles.

The EMG magnitude was quantiWed by computing themean EMG amplitude (%MVC) during performance. Wealso examined the frequency distribution of EMG signals toevaluate the level of muscle relaxation. Figure 2b showsexamples of the relative frequencies that are cumulativelyplotted as a function of the EMG activity level (%MVC).Here, the relative frequency at a given EMG activity levelindicates the relative duration in which EMG activity wasbelow the threshold level [muscle relaxation ratio (MRR),in % total performance duration]. Because approximately90% of the data points fell below 50% MVC for all themuscles, we examined the MRRs at the thresholds rangingfrom 5 to 50 with intervals of 5% MVC.

In addition, we estimated the co-contraction level ofantagonistic muscles in the forearm (i.e., ED and FD mus-cles) based on the relative diVerence signals (RDS) as pro-posed by Heuer (2007). The rectiWed and low-pass Wltered(6 Hz) EMG signals were scaled to a mean of 1, and for

Fig. 1 Typical examples of heart rate (HR, a) and sweat rate (SR, b)data. The graphs show the time series obtained during the Wrst 5 minof a participant’s performance (Beethoven: Piano Sonata No. 23 in Fminor, Op. 57, 1st Mov.). The left column shows the data recorded inthe rehearsal condition and the right column in the competition condi-tion

80

120

160

0

0.2

0.4

a: HR

Rehearsal Competition

bpm

mg/

min

/cm

2

1 min

b: SR

123

Exp Brain Res (2009) 199:117–126 121

each point in time the relative diVerence (E ¡ F)/(E + F)was computed, where E and F denote the scaled EMG val-ues of the ED and FD muscles, respectively (Fig. 2c).

According to Fujii et al. (2009b), we evaluated the co-con-traction level by calculating the standard deviation (SD) ofRDS for each piano performance [reciprocal contractionindex (RCI)]. A higher RCI value indicated the strongertendency toward reciprocal activity, whereas a lower RCIvalue indicated the stronger tendency toward co-contrac-tion of the antagonistic muscles.

Statistical analysis

We conducted paired t tests to examine the signiWcance ofdiVerences in the total performance score, scores for eachperformance evaluation item, VAS score, autonomic mea-sures, and RCI between the rehearsal and competition con-ditions. Because the data of mean EMG amplitudes did notfulWll the prerequisite for the application of parametricanalyses (i.e., normal distribution), Wilcoxon-matchedpairs signed-rank tests were used to determine the signiW-cance of the diVerences between conditions. As for the twoperformance subscores, we Wrst applied an one-way multi-variate ANOVA (MANOVA) to examine the eVect of con-dition, and then conducted paired t tests for the technicaland artistic scores separately. Regarding the MRR data, weWrst conducted a repeated-measures MANOVA with tworepeated factors: condition (2 levels: rehearsal and competi-tion) and threshold (10 levels: 5, 10, 15, 20, 25, 30, 35, 40,45, and 50% MVC). Then multiple two-way ANOVAswere further performed on the MRR data of each muscle.When the condition £ threshold interaction eVect wasfound to be signiWcant, we conducted one-way ANOVAsseparately for each of the 10 levels of threshold to deter-mine the simple main eVect of condition. P value of <0.05was regarded as statistically signiWcant.

Results

Performance quality

Table 1 shows the mean scores for each performance evalu-ation item in each condition. The stressful social-evaluativesituation led to impaired performance quality. The total per-formance score signiWcantly decreased from the rehearsalto competition condition [t(17) = 3.86, P < 0.01, Fig. 3a].The MANOVA performed on the data of technical andartistic scores revealed a signiWcant main eVect of condi-tion (Wilks’ lambda = 0.485, P < 0.01). The followingpaired t tests showed that both the technical and artisticscores were signiWcantly lower in the competition conditionthan in the rehearsal condition [technical: t(17) = 2.19,P < 0.05; artistic: t(17) = 4.25, P < 0.01, Fig. 3b]. Althoughthe diVerence in total performance duration between theconditions was statistically non-signiWcant (Wilcoxon

Fig. 2 Typical examples of electromyographic (EMG) data. The leftcolumn shows the data recorded in the rehearsal condition and the rightcolumn in the competition condition. The ED, FD, BB, and TR indicatethe extensor digitorum communis, Xexor digitorum superWcialis, bicepsbrachii, and upper trapezius muscles, respectively. a RectiWed andsmoothed EMG signals synchronized with sound data (Snd). The graphsshow the time series obtained during the Wrst 1,000 ms of a participant’sperformance (Messiaen: 8 Preludes pour Piano No.8 “Un ReXet dansle Vent”). b Histograms of EMG signals with cumulative frequencycurves. The histograms were produced based on the EMG signalsrecorded from the BB muscle while a participant was playing Liszt’sHungarian Rhapsody No. 2 in C-sharp minor. c Relative diVerence sig-nals (RDS) for antagonistic muscles in the forearm. The graphs show thetime series obtained during the Wrst 2,000 ms of a participant’s perfor-mance (ProkoWev: Piano Sonata No. 1 in F minor, Op. 1). The secondand third rows of graphs show the scaled (mean = 1) EMG signals of theED and FD muscles, respectively. The bottom row shows the RDS

a

c

Rehearsal Competition

ED

BB

TR

Snd

ED

FD

RDS

-10

0

10

0

20

40

0

40

80

0

60

0

60

-500

50

0

2

024

-1

0

1

Snd

FD

100 ms

0

20

40

60

80

100

30 200 10 20 40 50 0 10 30 40 50

EMG Activity Level (%MVC)

Rel

ativ

e F

requ

ency

(%

)

b

100 ms

%M

VC

%M

VC

%M

VC

%M

VC

mV

mV

n.u.

n.u.

n.u.

123

122 Exp Brain Res (2009) 199:117–126

Z = 1.02, P = 0.31), the duration was shorter in the compe-tition condition than in the rehearsal condition in 12 out of18 participants indicating that performing tempo tended toincrease in the competition condition.

Subjective state anxiety

The pre-performance VAS data showed that participantsreported a signiWcantly higher level of state anxiety in thecompetition condition than in the rehearsal condition[t(17) = 2.87, P < 0.05] conWrming the eVectiveness ofstress manipulation (Table 2).

Autonomic arousal

The psychological stress had remarkable eVects on auto-nomic arousal (Table 2). The mean HR during piano per-formance increased by a mean of 34.2 bpm from therehearsal to competition condition [t(17) = 9.47,P < 0.001]. The mean SR was also signiWcantly greater inthe competition condition [t(17) = 3.40, P < 0.01].

Muscular activity

We Wrst compared the mean EMG amplitude between thetwo conditions. The Wilcoxon signed-rank tests demon-strated that the mean EMG amplitude signiWcantlyincreased from the rehearsal to competition condition forthe BB (Wilcoxon Z = ¡3.10, P < 0.01) and TR (WilcoxonZ = ¡2.56, P < 0.05) muscles (Fig. 4a), while no signiWcantdiVerence was found for the other two muscles.

Then, we computed the MRRs at 10 selected thresholdsfor each muscle for each condition (Fig. 4b). TheMANOVA for MRR data revealed a signiWcant eVect ofcondition £ threshold interaction (Wilks’ lambda = 0.511,P < 0.05). ANOVA for each muscle further revealed thatonly the BB and TR muscles showed the signiWcant interac-tion eVects [BB: F(9,135) = 7.41, P < 0.001; TR:F(9,126) = 2.02, P < 0.05] and main eVects of condition[BB: F(1,15) = 6.17, P < 0.05; TR: F(1,14) = 5.63,P < 0.05]. As the interaction eVect was signiWcant for thesemuscles, we performed one-way ANOVAs at each thresh-old level to examine the simple main eVect of condition.The ANOVAs performed on the MRR data of BB muscleshowed that the MRRs at the thresholds of 5, 10, and 15%MVC were signiWcantly lower in the competition than inthe rehearsal condition (Ps < 0.05). The MRRs of TR mus-cle revealed a similar tendency, although the signiWcantdiVerences were found at slightly higher threshold levels,speciWcally at 15–50% MVC (Ps < 0.05).

Finally, we computed the RCI for the ED and FD mus-cles to examine their co-contraction level. As shown in

Fig. 3 Means of performance measures. a Total performance score.b Technical and artistic scores. Error bars represent between-partici-pant SD and asterisks denote the signiWcant diVerences betweenconditions (*P < 0.05, **P < 0.01)

Total

Tot

al P

erfo

rman

ce S

core

0

20

40

60

80

100

**

RehearsalCompetition

Technical Artistic

Mea

n P

erfo

rman

ce S

core

0

2

4

6

8

10

* **

a b

Table 2 Means (SDs) of the self-reported anxiety score and auto-nomic measures

*SigniWcance of the diVerence between conditionsa Visual analog mood scaleb Heart ratec Sweat rate

Rehearsal Competition P*

VASa score (mm) 44.4 (29.4) 56.1 (29.1) <0.05

Mean HRb (bpm) 112.4 (23.6) 146.6 (19.2) <0.001

Mean SRc (mg/min/cm2)

0.2295 (0.1540) 0.3195 (0.1245) <0.01

Table 1 Mean scores (SDs) for each performance evaluation item

Subcategory Item Rehearsal Competition P*

Technical Accuracy 6.4 (1.4) 6.1 (2.0) n.s.

Technical dexterity 6.4 (1.4) 5.9 (1.8) <0.10

Tempo and rhythm 6.4 (1.3) 5.9 (1.4) <0.05

Memory 7.5 (1.3) 6.9 (1.8) n.s.

Artistic Artistry 6.5 (1.5) 5.7 (1.8) <0.01

Interpretation 6.5 (1.5) 5.9 (1.6) <0.05

Expressiveness 6.4 (1.2) 5.8 (1.5) <0.05

Structural strength 6.5 (1.3) 6.0 (1.6) <0.05

Melodic and harmonic balance 6.8 (1.2) 5.6 (1.7) <0.001

Tone quality 6.5 (1.2) 5.5 (1.7) <0.01*SigniWcance of the diVerence between conditions

123

Exp Brain Res (2009) 199:117–126 123

Fig. 5, the RCI was signiWcantly lower in the competitioncondition than in the rehearsal condition [t(13) = 3.55,P < 0.01].

Discussion

The present experiment placed skilled pianists in a real,competitive, and extremely stressful performance situationto observe the changes in performance quality and insubjective, autonomic, and motor stress responses. Wefound that performance quality was considerably impairedin the competition condition. Despite the fact that many

musicians concern about performance impairment causedby MPA, only a few studies have directly compared musi-cians’ performance quality in a stressful condition with thatin a non-stressful condition. Most of such studies failed toWnd any signiWcant diVerences in performance qualitybetween evaluative and non-evaluative conditions (Craskeand Craig 1984; Yoshie et al. 2008a), and one study evenreported an improvement in performance quality from thenon-jury to jury condition (Hamann and Sobaje 1983).These inconsistent Wndings may partly be attributed to thediVerences in the induced stress levels. All of the threemajor theories concerning the arousal- or anxiety–perfor-mance relationship, i.e., inverted-U hypothesis (Yerkes andDodson 1908), ZOF model (Hanin 1978; Yoshie et al.2008b), and catastrophe theory (Hardy and ParWtt 1991)indicate that psychological stress beyond the optimal levelleads to performance decrements. Because the presentexperiment employed an audience larger than those used inprevious studies and authoritative judges including twoworld class pianists, the induced stress levels may havebeen high enough to adversely aVect performance quality.The observed changes in stress responses would explain onhow the social-evaluative situation impaired performancequality to such an extent.

The self-reported VAS score indicated that participantsfelt more anxious in the competition condition than in therehearsal condition. As the only diVerence between the twoconditions was the presence or absence of an audience andjudges, the increase in subjective anxiety appears to haveresulted from the misdirection of attention to task-irrelevantsocial cues. The relative shortage of attention paid to task-relevant cues (Wine 1971) might account for the fact thatthe decline in performance quality was greater in the artisticwhen compared with technical aspects, because artisticexpression requires high-order associative functionsinvolving the exploitation of real-time auditory feedbackand the broad temporal integration of musical elements.Moreover, psychosocial stress has been indicated to disruptfunctional connectivity within a frontoparietal network,

Fig. 4 Measures of EMG magnitude. a Box and whisker plots of themean EMG amplitude. The box represents the middle half of data, thehorizontal line in the box is the median, and the whisker is at 1.5 timesthe range of the middle half of data from the ends of the box. b Meansof muscle relaxation ratio (MRR) at the thresholds of 5–50% MVC.Error bars represent between-participant SD and asterisks denote thesigniWcant diVerences between conditions (*P < 0.05, **P < 0.01)

Mea

n E

MG

Am

plitu

de (

%M

VC

)

0

10

20

30

40

50

**

*RehearsalCompetition

ED FD BB TR

a

0

20

40

60

80

100

0

20

40

60

80

100

10 20 30 40 50

**

**

10 20 30 40 50

**

** * * * *

RehearsalCompetition

b ED FD

BB TR

Threshold (%MVC)

MR

R (

%P

erfo

rman

ce D

urat

ion)

Fig. 5 Means of reciprocal contraction index (RCI). Error barsrepresent between-participant SD and asterisks denote the signiWcantdiVerence between conditions (**P < 0.01)

RC

I

0.3

0.4

0.5

**

RehearsalCompetition

123

124 Exp Brain Res (2009) 199:117–126

deterring people from shifting attention (Liston et al. 2009).Therefore, the diYculty in shifting attention from task-irrelevant- to task-relevant cues may have further aggra-vated the subjective experience of anxiety in the presentparticipants.

The social-evaluative stressor aVected autonomicarousal as well. The HR during performance dramaticallyincreased from the rehearsal to competition condition,corroborating previous Wndings (Abel and Larkin 1990;Brotons 1994; Craske and Craig 1984; Fredrikson andGunnarsson 1992; LeBlanc et al. 1997; Yoshie et al.2008a). We could also duplicate the increase in SRobserved in a laboratory experiment (Yoshie et al. 2008a).The fact that our results showed greater changes in auto-nomic measures than those reported in studies using labora-tory-induced stress (Finsen et al. 2001; Matsumoto et al.2001; Noteboom et al. 2001; Wahlström et al. 2002; Yoshieet al. 2008a) suggests that the actual presence of a largeaudience and authoritative judges was an extreme socialstressor for a performing pianist. The present Wndings sug-gest that the stressful performance situation increased theactivation of sympathetic division of the autonomic ner-vous system (ANS). From an evolutionary perspective, thisstate is regarded as tuning the ANS to ensure metabolicresources for active contact defenses against attacks frompredators or conspeciWcs during the circa-strike stage(Fanselow 1994; Lang et al. 1997), and midbrain structuresincluding the periaqueductal gray seem to be responsiblefor integrating such fear responses (Mobbs et al. 2007).High levels of autonomic arousal, despite being an evolu-tionarily adaptive response in life-threatening situations,can have detrimental eVects in music performance becausethey not only fatigue musicians, but also alter their tempo-ral perception. Boltz (1994) suggested that an increase inarousal accelerates an individual’s inherent rate of activitycalled “internal tempo.” This might explain the reason whyperforming tempo tended to increase in the competitioncondition. Furthermore, among the four performance evalu-ation items in the technical category, only the score fortempo and rhythm signiWcantly decreased from therehearsal to competition condition (Table 1). Becauseappropriate timing is critical for optimal music perfor-mance, the stress-induced change in temporal perceptionmay have impaired participants’ performances.

In line with our hypothesis, the mean EMG amplitude ofthe BB and TR muscles signiWcantly increased, whereas theMRRs were reduced, from the rehearsal to competitioncondition. The results are consistent with the previous stud-ies showing increased EMG activity of the BB and/or TRmuscles during the execution of various motor tasks underpsychological stress (Au and Keir 2007; Bloemsaat et al.2005; van Galen et al. 2002; Visser et al. 2004; Wahlströmet al. 2002; Wahlström et al. 2003; Yoshie et al. 2008a).

It would be reasonable to consider that such elevated mus-cle activity in the upper extremity has reXected increasedcorticospinal motor tract excitability associated with nega-tive emotions (Baumgartner et al. 2007; Hajcak et al. 2007;Schutter et al. 2008). Although the underlying neurobio-logical mechanisms are still unclear, a possible explanationwould be that a fear-related amygdala excitation induced bythe psychosocial stressor activated neural circuits withinthe basal ganglia and thalamus that have the capacity toenhance motor cortex excitability (Coombes et al. 2008).Furthermore, the increased TR muscle activity in the com-petition condition is particularly of interest, because previ-ous studies have indicated the sensitivity of the TR muscleto psychosocial stimuli and the prevalence of shoulder painamong workers engaging in jobs characterized by lowphysical load and high psychosocial stress (Lundberg 2002;van Galen et al. 2002; Visser et al. 2004; Wahlström et al.2002; Wahlström et al. 2003). One plausible interpretationof these Wndings would be that TR muscle activity is moreclosely related to autonomic arousal when compared withother muscles. Alternatively, the TR muscle activationmight reXect compensatory motions of the shoulder to dealwith impaired manual dexterity under psychological stress.

As for the ED and FD muscles, however, the changingpatterns of EMG magnitude varied considerably amongparticipants, and no signiWcant group diVerence wasobserved between conditions. The results seemingly contra-dict previous studies that demonstrated signiWcant increasesin the EMG activity of these muscles under psychologicalstress (Finsen et al. 2001; Visser et al. 2004; Wahlströmet al. 2002; Yoshie et al. 2008a). Importantly, a recentexperiment using a repetitive tapping task (Bloemsaat et al.2005) showed that executive distal upper-limb muscles(including the ED and FD muscles) were less responsive tocognitive stress than proximal supporting muscles (includ-ing the BB and TR muscles). It is feasible, therefore, tospeculate that the eVects of social-evaluative stress on theforearm muscles were so subtle as to be masked by thediVerences in individual coping strategies to minimize per-formance decrements. This might explain why the muscu-loskeletal problems of the shoulder and neck are morefrequent (20–22%) compared with those of hand and fore-arm (7–16%) among musicians (Fishbein et al. 1988), sincethey are constantly put under competitive pressure.

Despite the absence of any signiWcant changes in EMGmagnitude, the ED and FD muscles revealed higher levelsof co-contraction in the competition condition, supportingboth our hypothesis and previous Wndings (Meulenbroeket al. 2005; van Galen et al. 2002; Weinberg and Hunt1976). The co-contraction of antagonistic muscles in theforearm has been suggested to be reduced as musiciansacquire expertise in piano performance (Furuya andKinoshita 2008) or in drumming performance (Fujii et al.

123

Exp Brain Res (2009) 199:117–126 125

2009a, b). Therefore, we could argue that the psychologicalstress led to a temporary regress to muscle activity patternsat an earlier stage of motor learning, which is characterizedby the freezing of excessive biomechanical degrees offreedom in an attempt to turn the motor apparatus into acontrollable state (Bernstein 1967; Pijpers et al. 2003).Heightening joint stiVness through co-contraction is gener-ally seen as one of the strategies to low-pass Wlter neuromo-tor signals in the face of deteriorated signal-to-noise ratio inthe motor system concomitant with anxiety (van Galenet al. 2002; van Gemmert and van Galen 1997), and co-contraction can actually contribute to the prevention ofperformance impairment under psychological stress in asimple motor task (Gribble et al. 2003; van Galen et al.2002; Yoshie et al. 2008a). Given that the present partici-pants showed signiWcant performance deterioration in thecompetition condition, however, the “co-contraction strat-egy” may not necessarily be eVective in maintaining thequality of piano performance, which is obviously the mostcomplex manual motor task ever used in stress studies.

As is the case with autonomic arousal, the higher levelsof proximal muscle activity and co-contraction of antago-nistic muscles in the forearm under social-evaluative stressare considered to be integral components of an adaptivebiological system and essential for adapting behaviors to anunstable risky environment. However, if these responsesoccur in a non-life-threatening piano performance, they canadversely inXuence the performing pianist. The increasedmuscle activity and joint stiVness may disrupt the subtlecontrol of loudness (Yoshie et al. 2008a) and the mainte-nance of temporal continuity (Drake and Palmer 2000),both of which are fundamental to musical expression.Moreover, muscle fatigue arising from the metabolicallydemanding hyperactivation and co-contraction can lead tothe decrease in the key-depression force production capac-ity (Furuya and Kinoshita 2008). The altered muscle activ-ity under psychological stress, if continued over a longperiod of time, can also add to the risk of playing-relatedmusculoskeletal disorders such as tenosynovitis and focaldystonia (Altenmüller and Jabusch 2009; Furuya et al.2006; Sakai 2002). The present study provides a new per-spective on the nature of MPA by considering pianists’responses to a social-evaluative performance situation as apart of defensive mechanisms. Future experiments shouldemploy a uniWed performing task and incorporate measure-ments of musical instrument digital interface signals toenable detailed note-by-note analyses of sound and physio-logical data. In addition, more studies are needed to exam-ine more in detail the relationships between the subjective,autonomic, and motor stress responses and the individualdiVerences in these responses to fully elucidate the factorsthat divide stress-tolerant musicians from their stress-intol-erant counterparts.

Acknowledgments This study was supported by the Grant-in-Aidfor ScientiWc Research (#19300216) of Japan Society for the Promo-tion of Science (JSPS) awarded to T. Ohtsuki and the Grant-in-Aid forJSPS Fellows (#2011133) awarded to M. Yoshie. We express our deepappreciation to Wve judges of the piano competition, namely Ms. ShokoSugitani, Dr. Hermann Gottschewski, Ms. Ikuko Endo, Mr. NobumichiKoyanagi, and Ms. Eriko Kanazawa, and to the staV and contestants ofthe competition.

References

Abel JL, Larkin KT (1990) Anticipation of performance among musi-cians: physiological arousal, conWdence, and state-anxiety.Psychol Music 18:171–182

Altenmüller E, Jabusch HC (2009) Focal hand dystonia in musicians:phenomenology, etiology, and psychological trigger factors.J Hand Ther 22:144–155

Au AK, Keir PJ (2007) Interfering eVects of multitasking on muscle activ-ity in the upper extremity. J Electromyogr Kinesiol 17:578–586

Baumgartner T, Willi M, Jäncke L (2007) Modulation of corticospinalactivity by strong emotions evoked by pictures and classical mu-sic: a transcranial magnetic stimulation study. NeuroReport18:261–265

Bernstein N (1967) The co-ordination and regulation of movement.Pergamon Press, New York

Bloemsaat JG, Meulenbroek RGJ, van Galen GP (2005) DiVerentialeVects of mental load on proximal and distal arm muscle activity.Exp Brain Res 167:622–634

Boltz MG (1994) Changes in internal tempo and eVects on the learningand remembering of event durations. J Exp Psychol Learn MemCogn 20:1154–1171

Brotons M (1994) EVects of performing conditions on music perfor-mance anxiety and performance quality. J Music Ther 31:63–81

Butler T, Pan H, Tuescher O, Engelien A, Goldstein M, Epstein J,Weisholtz D, Root JC, Protopopescu X, Cunningham-Bussel AC,Chang L, Xie XH, Chen Q, Phelps EA, Ledoux JE, Stern E,Silbersweig DA (2007) Human fear-related motor neurocircuitry.Neuroscience 150:1–7

Cannon WB (1915) Bodily changes in pain, hunger, fear, and rage.Appleton, New York

Cella DF, Perry SW (1986) Reliability and concurrent validity of threevisual-analog mood scales. Psychol Rep 59:827–833

Coombes SA, Cauraugh JH, Janelle CM (2006) Emotion and move-ment: activation of defensive circuitry alters the magnitude of asustained muscle contraction. Neurosci Lett 396:192–196

Coombes SA, Gamble KM, Cauraugh JH, Janelle CA (2008) Emo-tional states alter force control during a feedback occluded motortask. Emotion 8:104–113

Cox WJ, Kenardy J (1993) Performance anxiety, social phobia, andsetting eVects in instrumental music students. J Anxiety Disord7:49–60

Craske MG, Craig KD (1984) Musical performance anxiety: the three-systems model and self-eYcacy theory. Behav Res Ther 22:267–280

Drake C, Palmer C (2000) Skill acquisition in music performance: rela-tions between planning and temporal control. Cognition 74:1–32

Fanselow MS (1994) Neural organization of the defensive behaviorsystem responsible for fear. Psychon Bull Rev 1:429–438

Finsen L, Sogaard K, Jensen C, Borg V, Christensen H (2001) Muscleactivity and cardiovascular response during computer-mousework with and without memory demands. Ergonomics 44:1312–1329

Fishbein M, Middlestadt SE, Ottati V, Straus S, Ellis A (1988) Medicalproblems among ICSOM musicians: overview of a nationalsurvey. Med Probl Perform Art 3:1–8

123

126 Exp Brain Res (2009) 199:117–126

Fredrikson M, Gunnarsson R (1992) Psychobiology of stage fright: theeVect of public performance on neuroendocrine, cardiovascularand subjective reactions. Biol Psychol 33:51–61

Fujii S, Kudo K, Ohtsuki T, Oda S (2009a) Tapping performance andunderlying wrist muscle activity of non-drummers, drummers,and the world’s fastest drummer. Neurosci Lett 459:69–73

Fujii S, Kudo K, Shinya M, Ohtsuki T, Oda S (2009b) Wrist muscleactivity during rapid unimanual tapping with a drumstick indrummers and non-drummers. Motor Control 13:237–250

Furuya S, Kinoshita H (2008) Organization of the upper limb move-ment for piano key-depression diVers between expert pianists andnovice players. Exp Brain Res 185:581–593

Furuya S, Nakahara H, Aoki T, Kinoshita H (2006) Prevalence andcausal factors of playing-related musculoskeletal disorders of theupper extremity and trunk among Japanese pianists and pianostudents. Med Probl Perform Art 21:112–117

Gribble PL, Mullin LI, Cothros N, Mattar A (2003) Role of cocontrac-tion in arm movement accuracy. J Neurophysiol 89:2396–2405

Hajcak G, Molnar C, George MS, Bolger K, Koola J, Nahas Z (2007)Emotion facilitates action: a transcranial magnetic stimulationstudy of motor cortex excitability during picture viewing. Psycho-physiology 44:91–97

Hamann DL, Sobaje M (1983) Anxiety and the college musician: astudy of performance conditions and subject variables. PsycholMusic 11:37–50

Hanin YL (1978) A study of anxiety in sports. In: Straub WF (ed) Sportpsychology: an analysis of athlete behavior. Mouvement, Ithaca,pp 236–249

Hardy L, ParWtt G (1991) A catastrophe model of anxiety and perfor-mance. Br J Psychol 82:163–178

Heuer H (2007) Control of the dominant and nondominant hand:exploitation and taming of nonmuscular forces. Exp Brain Res178:363–373

Kudo K, Ohtsuki T (1998) Functional modiWcation of agonist–antago-nist electromyographic activity for rapid movement inhibition.Exp Brain Res 122:23–30

Lang PJ, Bradley MM, Cuthbert BN (1997) Motivated attention: aVect,activation, and action. In: Lang PJ, Simons RF, Balaban MT (eds)Attention and orienting: sensory and motivational processes.Erlbaum, Hillsdale, pp 97–135

LeBlanc A, Jin YC, Obert M, Siivola C (1997) EVect of audience onmusic performance anxiety. J Res Music Educ 45:480–496

Liston C, McEwen BS, Casey BJ (2009) Psychosocial stress reversiblydisrupts prefrontal processing and attentional control. Proc NatlAcad Sci USA 106:912–917

Lundberg U (2002) Psychophysiology of work: stress, gender, endo-crine response, and work-related upper extremity disorders. AmJ Ind Med 41:383–392

Matsumoto N, Yanase-Fujiwara M, Ohtsuki T (2001) EVects of trialimportance on the accurate ball-stroking movements. Proceed-ings of 18th Congress International Society of Biomechanics

Meulenbroek RGJ, van Galen GP, Hulstijn M, Hulstijn W, BloemsaatG (2005) Muscular co-contraction covaries with task load tocontrol the Xow of motion in Wne motor tasks. Biol Psychol68:331–352

Mobbs D, Petrovic P, Marchant JL, Hassabis D, Weiskopf N, SeymourB, Dolan RJ, Frith CD (2007) When fear is near: threat immi-nence elicits prefrontal-periaqueductal gray shifts in humans.Science 317:1079–1083

Noteboom JT, Fleshner M, Enoka RM (2001) Activation of the arousalresponse can impair performance on a simple motor task. J ApplPhysiol 91:821–831

Osu R, Franklin DW, Kato H, Gomi H, Domen K, Yoshioka T, KawatoM (2002) Short- and long-term changes in joint co-contractionassociated with motor learning as revealed from surface EMG.J Neurophysiol 88:991–1004

Papageorgi I, Hallam S, Welch GF (2007) A conceptual framework forunderstanding musical performance anxiety. Res Stud MusicEduc 28:83–107

Pijpers JR, Oudejans RRD, Holsheimer F, Bakker FC (2003) Anxiety-performance relationships in climbing: a process-orientedapproach. Psychol Sport Exerc 4:283–304

Sakai N (2002) Hand pain attributed to overuse among professionalpianists: a study of 200 cases. Med Probl Perform Art 17:178–180

Schutter D, Hofman D, Van Honk J (2008) Fearful faces selectivelyincrease corticospinal motor tract excitability: a transcranialmagnetic stimulation study. Psychophysiology 45:345–348

van Galen GP, Muller M, Meulenbroek RGJ, van Gemmert AWA(2002) Forearm EMG response activity during motor perfor-mance in individuals prone to increased stress reactivity. Am J IndMed 41:406–419

van Gemmert AWA, van Galen GP (1997) Stress, neuromotor noise,and human performance: a theoretical perspective. J Exp PsycholHum Percept Perform 23:1299–1313

Visser B, de Looze MP, de GraaV MP, van Dieën JH (2004) EVects ofprecision demands and mental pressure on muscle activation andhand forces in computer mouse tasks. Ergonomics 47:202–217

Wahlström J, Hagberg M, Johnson PW, Svensson J, Rempel D (2002)InXuence of time pressure and verbal provocation on physiologi-cal and psychological reactions during work with a computermouse. Eur J Appl Physiol 87:257–263

Wahlström J, Lindegard A, Ahlborg G, Ekman A, Hagberg M (2003)Perceived muscular tension, emotional stress, psychologicaldemands and physical load during VDU work. Int Arch OccupEnviron Health 76:584–590

Weinberg RS, Hunt VV (1976) Interrelationships between anxiety,motor performance and electromyography. J Mot Behav 8:219–224

Wilson GD (1997) Performance anxiety. In: Hargreaves DJ, North AC(eds) The social psychology of music. Oxford University Press,Oxford, pp 229–245

Wine J (1971) Test anxiety and direction of attention. Psychol Bull76:92–104

Yerkes RM, Dodson JD (1908) The relationship of strength of stimulusto rapidity of habit formation. J Comp Neurol Psychol 18:459–482

Yoshie M, Kudo K, Ohtsuki T (2008a) EVects of psychological stresson state anxiety, electromyographic activity, and arpeggio perfor-mance in pianists. Med Probl Perform Art 23:120–132

Yoshie M, Shigemasu K, Kudo K, Ohtsuki T (2008b) Multidimen-sional anxiety and music performance: an exploratory applicationof the zones of optimal functioning model. In: Buchwald P,Ringeisen T, Eysenck M (eds) Stress and anxiety: application tolife span development and health promotion. Logos Verlag,Berlin, pp 163–171

Yoshie M, Shigemasu K, Kudo K, Ohtsuki T (2009) EVects of stateanxiety on music performance: relationship between the RevisedCompetitive State Anxiety Inventory-2 subscales and pianoperformance. Music Sci 13:55–84

123